Clays and Clay Minerals, Vol. 43, No. 2, 212-222, 1995. EXPERIMENTAL STUDY ON THE FORMATION OF CLAY MINERALS FROM OBSIDIAN BY INTERACTION WITH ACID SOLUTION AT 150" AND 200*C MOTOHARU KAWANO 1 AND KATSUTOSHI TOMITA 2 ' Department of Environmental Sciences and Technology, Faculty of Agriculture, Kagoshima University 1-21-24 Korimoto, Kagoshima 890, Japan Institute of Earth Sciences, Faculty of Science, Kagoshima University 1-21-35 Korimoto, Kagoshima 890, Japan Abstract--Experimental alteration of obsidian with HC1 solution was performed to elucidate dissolution mechanism and formation process of clay minerals in acid solution. Reactions were carried out using 0.1, 0.5, and 4.0 g of obsidian to 100 ml of 0.01 N HCI solution at 150" and 200"C for 1 to 60 days. The reaction products were examined by X-ray powder diffraction, scanningelectron microscopy, transmission electron microscopy (TEM), and energy dispersive X-ray analysis. The surface composition of obsidian before and after alteration was investigated by X-ray photoelectron spectroscopy (XPS). TEM showed that boehmite precipitated at early stage and spherical kaolinite appeared subsequently by 200*(2reactions. However, spherical halloysite occurred predominantly with small amounts of allophane, boehmite, and kaolinite by 150*(3reaction in which formation process of the halloysite from allophane passing through an intermediate phase of small size rounded aggregate that consists of fine particles of allophane was observed. A boehmite exhibiting hexagonal platy habit with higher degree of crystallinity was formed by 200"C reaction as a stable phase in solution containing lower Si concentration at which the solution composition coincides with the stability field of boehmite on activity diagram for the system Na~O- m1203-SiO2-H20. The fibrous boehmite having lower crystallinity appeared with increasing Si concen- tration, considered as a metastable phase in the stability field of kaolinite. XPS indicated that dissolution of obsidian in acid solution proceeded initiallyby cation exchange between Na ions and hydronium ions in solution and subsequently by preferential release of Al ions relative to Si from the Na depleted surface. Key Words--Acid solution, Allophane, Boehmite, Experimental alteration, Halloysite, Obsidian, Spher- ical kaolinite. INTRODUCTION Clay minerals are widely occurred by interface re- action between the earth's surface constituents and aqueous solution in which the minerals tend to trans- form very slowly into more stable secondary minerals. Allophane, halloysite, and kaolinite are common clay minerals distributed extensively on the earth's surface, and it has been presumed that allophane is formed at first and transformed subsequently into halloysite and then finally into kaolinite (Tamura and Jackson 1953, Sudo and Takahashi 1956, Siefferann and Millot 1969, Sudo and Takahashi 1971, Thomas 1971, La Iglesia and Galan 1975, Nagasawa 1978a, 1978b). The for- mation conditions of these materials have been exten- sively discussed from the view point of equilibrium relations on activity diagram and synthesis of these materials has been conducted using the equilibrium relations (Robbie and Waldbaum 1968, Helgeson et al 1969, Kittrick 1970, Helgeson 1971, Huang and Keller 1973, La Iglesia et al 1976, La Iglesia and Van Oos- terwyck-Gastuche 1978, Tsuzuki and Suzuki 1980, Tsuzuki and Kawabe 1983). On the other hand, various clay minerals including kaolinite and haUoysite have been also synthesized from a view point of alteration of starting materials. Gruner Copyright 1995, The Clay Minerals Society (1944), in his experimental work on feldspar alteration, found that kaolinite and pyrophyllite were formed from albite in HC1 solution at 300~ Morey and Chen (1955) performed experimental alteration of albite in water, and reported that albite was altered to boehmite and kaolinite at 200~ and analcime, boehmite, K-mica, and paragonite at 300~ Similarly, Brindley and Ra- doslovich (1956) reported that boehmite and kaolinite appeared as main alteration products of albite in 0.1 N HC1 solutions at 285 ~ to 435~ Morey and Fournier (1961) identified boehmite, paragonite, and amor- phous material at 295~ Parham (1969) likewise ob- tained halloysite and boehmite from K-feldspar and plagioclase, respectively, by reaction with water at 77 ~ 78"C for 140 days. Tsuzuki and Suzuki (1980) carried out experimental alteration of labradorite in HC1 so- lution at 230~ and reported that amorphous silica and boehmite were initially formed and later kaolinite appeared. At room temperature condition, Busenberg (1978) obtained gibbsite and microcrystalline halloy- site as alteration products ofalbite in aqueous solution. However, dissolution process of volcanic glass in acid solutions has not been well understood and formation process of the reaction products has not been suffi- ciently investigated. The present study performed experimental alter- 212
Transcript
Clays and Clay Minerals, Vol. 43, No. 2, 212-222, 1995.
EXPERIMENTAL STUDY ON THE FORMATION OF CLAY MINERALS FROM OBSIDIAN BY INTERACTION WITH ACID
SOLUTION AT 150" AND 200*C
MOTOHARU KAWANO 1 AND KATSUTOSHI TOMITA 2
' Department of Environmental Sciences and Technology, Faculty of Agriculture, Kagoshima University 1-21-24 Korimoto, Kagoshima 890, Japan
Institute of Earth Sciences, Faculty of Science, Kagoshima University 1-21-35 Korimoto, Kagoshima 890, Japan
Abstract--Experimental alteration of obsidian with HC1 solution was performed to elucidate dissolution mechanism and formation process of clay minerals in acid solution. Reactions were carried out using 0.1, 0.5, and 4.0 g of obsidian to 100 ml of 0.01 N HCI solution at 150" and 200"C for 1 to 60 days. The reaction products were examined by X-ray powder diffraction, scanning electron microscopy, transmission electron microscopy (TEM), and energy dispersive X-ray analysis. The surface composition of obsidian before and after alteration was investigated by X-ray photoelectron spectroscopy (XPS). TEM showed that boehmite precipitated at early stage and spherical kaolinite appeared subsequently by 200*(2 reactions. However, spherical halloysite occurred predominantly with small amounts of allophane, boehmite, and kaolinite by 150*(3 reaction in which formation process of the halloysite from allophane passing through an intermediate phase of small size rounded aggregate that consists of fine particles of allophane was observed. A boehmite exhibiting hexagonal platy habit with higher degree of crystallinity was formed by 200"C reaction as a stable phase in solution containing lower Si concentration at which the solution composition coincides with the stability field of boehmite on activity diagram for the system Na~O- m1203-SiO2-H20. The fibrous boehmite having lower crystallinity appeared with increasing Si concen- tration, considered as a metastable phase in the stability field of kaolinite. XPS indicated that dissolution of obsidian in acid solution proceeded initially by cation exchange between Na ions and hydronium ions in solution and subsequently by preferential release of Al ions relative to Si from the Na depleted surface. Key Words--Acid solution, Allophane, Boehmite, Experimental alteration, Halloysite, Obsidian, Spher- ical kaolinite.
I N T R O D U C T I O N
Clay minerals are widely occurred by interface re- action between the earth's surface constituents and aqueous solution in which the minerals tend to trans- form very slowly into more stable secondary minerals. Allophane, halloysite, and kaolinite are common clay minerals distributed extensively on the earth's surface, and it has been presumed that allophane is formed at first and transformed subsequently into halloysite and then finally into kaolinite (Tamura and Jackson 1953, Sudo and Takahashi 1956, Siefferann and Millot 1969, Sudo and Takahashi 1971, Thomas 1971, La Iglesia and Galan 1975, Nagasawa 1978a, 1978b). The for- mation conditions of these materials have been exten- sively discussed from the view point of equilibrium relations on activity diagram and synthesis of these materials has been conducted using the equilibrium relations (Robbie and Waldbaum 1968, Helgeson et a l 1969, Kittrick 1970, Helgeson 1971, Huang and Keller 1973, La Iglesia et al 1976, La Iglesia and Van Oos- terwyck-Gastuche 1978, Tsuzuki and Suzuki 1980, Tsuzuki and Kawabe 1983).
On the other hand, various clay minerals including kaolinite and haUoysite have been also synthesized from a view point of alteration of starting materials. Gruner
Copyright �9 1995, The Clay Minerals Society
(1944), in his experimental work on feldspar alteration, found that kaolinite and pyrophyllite were formed from albite in HC1 solution at 300~ Morey and Chen (1955) performed experimental alteration of albite in water, and reported that albite was altered to boehmite and kaolinite at 200~ and analcime, boehmite, K-mica, and paragonite at 300~ Similarly, Brindley and Ra- doslovich (1956) reported that boehmite and kaolinite appeared as main alteration products of albite in 0.1 N HC1 solutions at 285 ~ to 435~ Morey and Fournier (1961) identified boehmite, paragonite, and amor- phous material at 295~ Parham (1969) likewise ob- tained halloysite and boehmite from K-feldspar and plagioclase, respectively, by reaction with water at 77 ~ 78"C for 140 days. Tsuzuki and Suzuki (1980) carried out experimental alteration of labradorite in HC1 so- lution at 230~ and reported that amorphous silica and boehmite were initially formed and later kaolinite appeared. At room temperature condition, Busenberg (1978) obtained gibbsite and microcrystalline halloy- site as alteration products ofalbite in aqueous solution. However, dissolution process of volcanic glass in acid solutions has not been well understood and formation process of the reaction products has not been suffi- ciently investigated.
The present study performed experimental alter-
212
Vol. 43, No. 2, 1995 Formation of clay minerals from obsidian in acid solution 213
ation o f obsidian with HC1 solution in closed system at 150* and 200"C, and dissolution process of the ob- sidian in acid solution and formation of reaction prod- ucts were examined.
M A T E R I A L A N D EXPERIMENTAL METHODS
An obsidian collected from Mifune, Kagoshima Pre- fecture, Japan is the material used in this study. Sam- pies were crushed, then particles o f obsidian between 50 to 100 mesh size were selected and cleaned ultra- sonically in aceton to remove adhering fine particles, and were used as a starting material. The chemical composit ion o f obsidian was first determined by elec- tron probe microanalyzer as: 80.04% SiO2, 12.27% A1203, 0.16% TiO2, 0.84% FeO, 0.18% MgO, 1.10% CaO, 3.14% Na20, and 3.04% K20. The experiments were performed at 150* and 200"C for 1 to 60 days using a Teflon bottle in which various amounts o f start- ing material and 100 ml of 0.01 N HCI solution were placed. The amounts of starting material used were 0.1, 0.5, and 4.0 g for 200~ experiments, and 0.5 and 4.0 g for 150"(2. The pressure was held at its equilib- r ium vapor pressure at each temperature. After differ- ent reaction times, the Teflon bottle was quenched and the reaction products, residual starting materials, and solutions were separated. The reaction products were examined by X-ray powder diffraction (XRD), scan- ning electron microscopy (SEM), transmission electron microscopy (TEM), and energy dispersive X-ray anal- ysis (EDX). The residual starting materials were in- vestigated by SEM and X-ray photoelectron spectros- copy (XPS). The concentrations of dissolved elements and pH values o f the solutions were measured. Details of the analytical techniques have been described by Kawano and Tomita (1992) and Kawano et al (1993).
RESULTS
Products of 200~C experiments
The experimental conditions, pH values of final so- lutions, and reaction products identified by X R D and TEM are summarized in Table 1. X R D profiles of the products formed by 200~ reactions indicated that boehmite appeared at early reaction stage and kaolinite was formed subsequently (Figure 1). For the reactions using 4.0 g of starting material, broad peaks corre- sponding to 020 (d = 6.15 A) and 120 (d = 3.18 A) reflections o fboehmi te were detected by XRD in 1 -day product. Small peaks due to basal reflections of ka- olinite appeared by 5-day reaction, which continuously increased in intensity with time. The X R D profile of kaolinite in 60-day product exhibited sharp basal re- flections at 7.19 and 3.58 A with poorly separated 11T and 1 T 0 reflections, suggesting highly disordered struc- ture (Hinckley 1963). For the reactions using 0.5 g of starting material, sharp 020 (d = 6.13 A), 120 (d = 3.17 A), and 031 (d = 2.352 A) reflections o fboehmi te
(A) 4.0g [
K 5 days
"2 # CuK a RADIATION
m) o.sg ! I (c) 0.1g -
, / ..... f ...... ] l[ I iAoo.,.I L ,I . . . . . i
"2#CuKr RADIATION "2#CuKa RADIATION
Figure 1. X-ray powder diffraction profiles of reaction prod- ucts formed by 200"C reactions using (a) 4.0, (b) 0.5, and (c) 0.1 g of starting material. The labels B and K represent dif- fraction peaks of boehmite and kaolinite, respectively. The labels 1 to 60 days signify reaction time.
were clearly observed in 3-day product, and the inten- sity gradually decreased with time. A small peak of 001 reflection of kaolinite appeared by 10-day reaction and then the basal reflections and broad non-basal reflec- tions increased in intensity with time. No significant diffraction peak could be detected in l -day product produced by reaction using 0.1 g of starting material. A sharp 020 (d = 6.12 J~) reflection of boehmite how- ever appeared after 3 days and small peaks o f basal reflections of kaolinite were found in 60-day product. Consequently, kaolinite tended to appear at early stage with large volume as amount of starting material were increased.
SEM of the reaction products of 4.0 g of starting material showed that the boehmite of 1-day reaction exhibited aggregated texture of flakes (Figure 2a). EDX indicated that the boehmite consisted mainly of Al and a small amount o f Si (Figure 2d). TEM confirmed that the flakes were composed of aggregates o f thin fibers elongated in various direction and of bundles of fibers (not shown). Boehmites exhibiting similar habi t have been reported by Tchoubar (1965) and Kawano and Tomita (1993). SEM of 5-day product showed that thin coating less than 0.1 pm thickness was precipitated on the obsidian surface on which very small spherical ka- olinite less than 0.1 pm in diameter and flaky boehmite were likewise observed (Figure 2b). EDX of the coating adhering very small amounts of both kaolinite and boehmite particles displayed strong peak o f AI and relatively small peak of Si (Figure 2e), indicating that the coating consisted ofaluminosi l icate of which chem- ical composit ion was rich in Al rather than Si. The coating disappeared as reaction proceeded and size of the spherical kaolinite increased with t ime (Figure 2c); average values o f the diameter are: 0.09, 0.13, 0.17, 0.24, 0.33, and 0.38 pm for 1- to 60-day products.
214 Kawano and Tomita Clays and Clay Minerals
Table 1. Experimental conditions and reaction products from obsidian by interaction with HC1 solution.
Boehmite(p) = hexagonal platy boehmite, Boehmite(l) = lath-shaped boehmite, Boehmite(t) = fibrous boehmite, Disk = rounded platy material exhibiting disk-like habit. ( ) = trace products detected by transmission electron microscopy.
pH-values of final solutions measured at room temperature. : The reaction products were identified by X-ray powder diffraction and transmission electron microscopy.
EDX (Figure 2t) confirmed that Al/Si ratio of the spheres is close to theoretical value of kaolinite (Al/Si = 1.0).
Using 0.5 g of starting material, the reaction at early stage produced boehmite that exhibits lath-like habit with hexagonal edge (Figure 3a). This however began to dissolve and new reaction phases of fibrous boehm- ite and spherical kaolinite appeared subsequently for longer duration at a later stage. In 1- and 3-day prod- ucts, small amount of rounded platy material exhib- iting disk-like habit with 0.5 to 1.5 #m in diameter was observed (Figure 3b). A clear rounded outline of the material was displayed but it later changed to dissolved habit with irregular and vague outline suggesting that the material tended to decompose as reaction pro- ceeded. Kawano and Tomita (1993) synthesized a sim- ilar rounded material by hydrothermal reaction be- tween obsidian and AIC13 solution at 200"C but the material was not characterized sufficiently and also re- mained unidentified. EDX spectrum of the rounded material showed strong X-ray peak of A1 and small peak of Si indicating that the material consisted mainly of A1 (Figure 3c). The electron diffraction in the direc-
tion normal to the surface of the rounded platy crystal showed a two-dimensional hexagonal net pattern (Fig- ure 3D). The reciprocal cell axes were assumed to be a* and c* as directions labeled on Figure 3D. Cell di- mensions determined from the diffraction data cali- brated by ring diffraction of gold are approximately a = 4.5 and c = 2.7/~. The values are close to the cell dimensions of diaspore (a = 4.396 and c = 2.844/~; Brown 1980) rather than those ofboehmite (a = 3.6936 and c = 2.8679/~; Christoph et al 1979). It is also well known that a luminum hydroxides occurring in the or- der as increasing the temperature are as follows: gibbs- ire (3'-Al(OH)3) ~ boehmite (3,-AIOOH) ~ diaspore (a-AIOOH). Roy and Osborn (1954) reported that boehmite-diaspore t ransformat ion temperature is 275~ indicating that formation temperature of dia- spore is considerably higher than that of present study (200~ However, we conclude that the rounded platy material may probably be diaspore on the basis of the EDX and electron diffraction data.
For the reaction using 0.1 g of starting material, hex- agonal platy boehmite ranging in size from about 0.5
Vol. 43, No. 2, 1995 Formation of clay minerals from obsidian in acid solution 215
to 5.0 #m was mainly formed at early stage in about 3 days. Small amounts of rounded plates less than 1.5 #m in diameter (Figure 4a) were also formed at this stage. It can be noted that aluminosilicate coating that precipitated on the obsidian surface at reaction using 4.0 g of starting material could not be observed in this 0. l g reaction. Both boehmite and rounded material exhibited clear outline which apparently began to dis- solve at early stage reaction and new phases of fibrous boehmite and spherical kaolinite appeared as reaction proceeded (Figure 4B). The hexagonal boehmite gave characteristically very well shaped XRD reflections as shown in Figure 1, and the d-values tended to decrease slightly as the morphology changed from fibrous to lath-shaped and to hexagonal habits.
Products of 150~ experiments
XRD profiles of the reaction products at 150~ gave no significant diffraction peak except for small angle scattering. However, small amounts of products as summarized in Table 1 were clearly observed by TEM. The reaction products using 4.0 g of starting material were predominated by boehmite and halloysite with small amounts of allophane and spherical kaolinite. The boehmite exhibited aggregated texture of circled or straight fibers less than about 2.0 #m in length sim- ilarly to the material as shown in Figure 2a. Some straight fibers were arranged in parallel orientation in bundles. The allophane appeared as two distinct tex- tures exhibiting homogeneous and heterogeneous ag- gregates, respectively (Figures 5a and 5b). The ho- mogeneous aggregate is the predominant form occur- ring in this system, which is composed of small par- tides of aUophane gathered homogeneously (Figure 5a). Such homogeneous aggregate is also a characteristic texture of natural allophane occurring in weathering environment (Wada 1989). The heterogeneous aggre- gate is formed by aggregation of small rounded grains less than 0.2 #m in diameter, which is possibly com- posed of allophane particles or intermediate phase be- tween allophane and spherical halloysite (Figure 5b). The halloysite mainly occurred in spherical form less than about 0.2 #m in diameter together with small amount of tubular habit (Figure 5c). These materials were produced by 1-day reaction and preserved well up to 60 days. A very small amount of spherical ka- olinite also appeared as a new reaction phase after five days and increased in size with time.
In the reaction products using 0.5 g of starting ma- terial, lath-shaped boehmite, rounded platy diaspor, and a small amount of allophane were produced at early reaction stage. Fibrous boehmite however ap- peared after 10 days.
Surface alteration in acid solution
In the course of dissolution, small rounded and tiny elongated etch pits appeared on the obsidian surface
Figure 2. Scanning electron micrographs of reaction prod- ucts formed by 200~ reactions using 4.0 g of starting material for (a) l day, (b) 3 days, and (c) 10 days. Also, energy dis- persive X-ray spectra (d, e, and f) obtained from the products of a, b, and c, respectively. The reaction products observed in these photographs are (a) fibrous boehmite, (b) fibrous boehmite and very small spherical kaolinite adhered to non- crystalline aluminosilicate coating precipitated on the obsi- dian surface, and (c) spherical kaolinite.
during reaction within 1 day and tended to develop in size and depth with time. The formation of etch pits indicates selective dissolution at specific sites of excess energy, such as dislocations, defects, and microcracks produced by sample preparation. Moreover, TEM showed thin leached layer less than 0. i ~m in thickness on the obsidian surface and the layer was noncrystalline for electron diffraction. The leached layer was com- posed of two distinct inner and outer layers less than about 0.08 and 0.03 ~m in thickness, respectively, and they were parallel to the obsidian surface with a sharp boundary and partly peeled off the surface.
The surface compositions of obsidian before and af- ter alteration at 150~ were examined by XPS. The obsidian samples after alteration at 200~ were not used for XPS analysis because large amounts of ad- hering kaolinite crystals remained on the obsidian sur- face even after ultrasonic cleaning. Figures 6a and 6b respectively show variations with NaKLL/Si2p and Al2p/ Si2v intensity ratios of obsidians altered by 150~ re-
216 Kawano and Tomita Clays and Clay Minerals
Figure 3. Transmission electron micrographs of reaction products formed by 200"C reactions using 0.5 g of starting material. (a) lath-shaped boehmite formed by 1-day reaction, (b) rounded material showing disk-like habit formed by 5-day reaction, (c) EDX spectrum of the rounded material, (d) electron diffraction of the rounded material b.
actions as a function of square root of time. The Nat.EL/ Si2p ratio of the original obsidian was 0.785 which decreased rapidly at the early reaction stage and still remained constant after. Similarly, the A12p/Si2p ratio of the original obsidian was 0.153 which decreased gradually as reaction proceeded. These observations suggest that initially Na ions were released from the obsidian surface probably by cation exchange reaction, and subsequently A1 ions were preferentially released relative to Si from the surface.
Chemistry of solutions
Figure 8 shows concentrations of dissolved Si, A1, and Na in solutions during dissolution process of ob- sidian. The concentrations of Na increased rapidly at early reaction stage and the increasing rates decreased
successively in each experiment. The leaching behav- iors of elements during dissolution process of glass and silicate minerals are well known as a parabolic leaching kinetic which is due to the formation of a protective layer on the surface (Chou and Wollast 1984) or to the preferential dissolution of fine particles and of sites with excess energy (Holdren and Berner 1979). The concentrations of Si and A1 as well as Na increased at early stage. However, the values varied depending on amount of starting material and on temperature after about 10 days. As shown in Table 1, reaction products consisting of Si and A1 were produced in all experi- ments. Thus, the released Si and A1 were consumed as components of the products so the concentrations of Si and AI were affected by amounts of the products. Other elements such as K, Ca, and Mg exhibited the
Vol. 43, No. 2, 1995 Formation of clay minerals from obsidian in acid solution 217
Figure 4. Scanning electron micrographs of products formed by 200~ reactions using 0.1 g of starting material. (a) hexagonal boehmite crystals and rounded materials formed by 1-day reaction, (b) fibrous boehmite formed by 30-day reaction. Arrows on Figure a indicate rounded materials.
same leaching behaviors as that of Na. However the values of concentrations appeared to be below at 50 ppm for K, 30 ppm for Ca, and 3 ppm for Mg through- out the 200"C reaction, and at 12 ppm for K, 10 ppm for Ca, and 0.5 ppm for Mg throughout the 150"C reaction.
DISCUSSION
In the reaction processes of obsidian with 0.01 N HC1 solution, boehmite precipitated at first and ka- olinite formed later at 200"C reactions, and halloysite appeared instead of kaolinite at 150~ reactions. The elements composed of obsidian structure are simul- taneously released into the solution following the dif- fusion control process, and reaction products would precipitate from the solution when its compositions correspond to their stability fields. The solution com- positions during the reactions at 200* and 150~ are plotted on activity diagrams for the system Na20-A1203- SiO2-H20) as shown in Figures 8A and 8B, respec- tively. For the 200"C reactions using 4.0 g of starting material, the solution compositions fall into the sta- bility field of kaolinite throughout the reactions. Then, kaolinite begins to precipitate in equilibrium with the soIution as a stable phase and grows progressively dur- ing the reaction. However, a small amount of fibrous
boehmite appeared at early reaction stage, which sug- gests that the boehmite is formed as a metastable phase. The solution compositions of reactions using 0.5 and 0. l g of starting materials correspond to stability field of boehmite at first and move to kaolinite field as re- action proceeds. The paths are consistent with the ap- pearance ofboehmite and kaolinite during the reaction as summarized in Table I.
The solution compositions of 150~ reactions using 4.0 g of starting material fall into the stability field of kaolinite, and those using 0.5 g of starting material move from the stability field of boehmite to that of kaolinite during the reactions (Figure 8b). As shown in Table 1, boehmite appeared as a most abundant product during both the series of reactions, however haUoysite occurred predominantly rather than kaolin- ite throughout the reactions using 4.0 g of starting ma- terial. Either kaolinite or halloysite could not be ob- served in products of the reactions using 0.5 g of start- ing material.
A small amount ofallophane was found in products of all experiments independent of the solution com- positions, suggesting that the material occurred as a metastable phase. Allophane was defined as a group name of naturally occurring hydrous aluminosilicates having short-range ordered structure (van Olphen
218 Kawano and Tomita Clays and Clay Minerals
~
or)
. . , I v
t~ z
13,,. N
o,.. m
0.8.
0.6
0.4
0.2
0.0
(A) o 4.09
�9 0.5 9
\ ~ ~ o "0~0~
I I I
0 2 4 6 D U R A T I O N (DAY 1/2)
0 . 2 o 4 o g
�9 0.5g
O
0"~ 0 - 0 ~ 0 ~ O ~ 0
0.0 I I I
8
0 2 4 6 8 D U R A T I O N (DAY 1/2)
Figure 6. Variations of A) Nar.LL/Si2p and B) Alzp/Si2v in- tensity ratios of X-ray photoelectron spectra of obsidian be- fore and after alteration at 150*(] as a function of square root of time. Stars indicate the values of original obsidian.
1971). The mater ia l consists o f ident ical spherical par- ticles as small as about 50 • in d iamete r (Watanabe and Sudo 1969, Hall et a! 1985), and the Si /Al ratios vary widely f rom 0.5 to 1.0 (Wada 1989). In nature,
Figure 5. Transmission electron micrographs of reaction products formed by 150"C reactions using 4.0 g of starting material. (a) homogeneous aggregate of allophane particles
4----
appeared by 10-day reaction, (b) heterogeneous aggregate of allophane particles that appeared in I 0-day reaction, (c) spher- ical halloysite formed by 5-day reaction. Arrows on Figure b indicate some of rounded grains composed of allophane par- tides, which are intermediate transitional forms from allo- phane to spherical halloysite.
Vol. 43, No. 2, 1995 Formation of clay minerals from obsidian in acid solution 219 ,oo 4f/ ],,or 400 �9 0,5 g o..,..~/ ~ o ~
0 20 40 60 0 20 40 60 0 20 40 60 D U R A T I O N (DAY) DURAT ION (DAY) DURATION (DAY)
Figure 7. Concentrations of dissolved Si, AI, and Na in solutions during dissolution process of obsidian. Upper = reaction at 200~ Lower = reaction at 150~
allophane is extensively formed under various envi- ronments on the earth's surface, for example, by weath- ering from volcanic glass (Allen and Haj ek 1989, Parfitt and Kimble 1989, Wada 1989) and feldspar (Tazaki 1978), and by direct precipitation in fiver stream (WeUs et a! 1977). For syntheses, allophane is readily pro- duced in acid solutions containing various proport ions of A1 and Si ions by heating at 95 ~ to 100~ for 113 h (Wada et a! 1979), and by acidifying the alkaline so- lutions containing the same cations (Farmer et al 1979, Wada and Wada 1981). These facts suggest that allo- phane is easily and rapidly formed in solutions having wide range of both A1/Si ratio and pH. Generally, ma- terials having higher rate of precipitat ion tend to be produced as a metastable phase in a supersaturated solution in which the rate of dissolution competes with that of precipitation. Therefore, allophane that ap- peared in the products of these reactions may be formed as a metastable phase.
Boehmite appeared all throughout the experiments. It exhibited hexagonal platy, lath-shaped, and fibrous habits depending on the reaction conditions. The hex- agonal platy boehmite was formed by 200~ reaction
as a stable phase in solution containing lower Si con- centration which coincides with the stabili ty field on activity diagram for the system Na20-kl203-SiO2-H20 (Figure 8). The lath-shaped and fibrous habits appeared successively as a metastable phase with increasing Si concentration which corresponds with stability field of kaolinite. For the 150~ reaction, lath-shaped boehm- ite was produced in solution corresponding with sta- bility field o fboehmi te but fibrous boehmite appeared as a metastable phase in the stability field o f kaolinite. The relationship between morphology ofboehmi te and formation conditions are schematically represented in Figure 9. It is well known that boehmite exhibits var- ious degrees of structural disorder (Papre et al 1958). The platy boehmites gave symmetrical shape reflec- tions of XRD profiles indicating higher crystallinity, whereas that of the fibrous boehmite was uniformly broad which suggests that the fibrous boehmite has highly disordered structure (Figure 1). Furthermore, the d-values of reflections tended to increase with in- creasing structural disorder. Calvet et al (1953) used the term pseudoboehmite for such a disordered boehm- ite. Later, Tettenhorst and Hofmann (1980) studied
'1-
s
(B)
Z
(.'3 0 ._1
4
3
2
1
0
-1
-2 -5
(A)
Boehmite
4.0g
0.5g
0.1g I { 2 t
- 4 -3
~ Smectite
[] ?~ - 2 -1
LOG[H4S iO41
7" -r
§
z 0 0 .J
4
3
2
1
0
-1
-2 -5
Boehmite
o 4.0g
�9 0.Sg 0 ~ QI,
- 4 - 3
~ 1
- 2 -1
220 Kawano and Tomita Clays and Clay Minerals
LOG[H4S i04 ]
Figure 8. Plots of solution compositions on the activity dia- grams for the system of Na:O-AI203-SiO2-H20 at (a) 200* and (b) 150~ The positions of stability field boundaries were derived from Helgeson (1971 ).
crystal chemistry of boehmites having various degrees of disorder, and pointed out that pseudoboehmite is essentially finely crystalline boehmite containing more sorbed water intercalated randomly between octahe- dral layers.
Kaolinite was mainly formed by 200~ reactions but halloysite appeared predominant ly by 150~ reaction with trace amount of kaolinite. The kaolinite exhibited characteristic spherical form similarly to that synthe- sized by Tomura et al(1983, 1985), and no platy form
Hexagonal ..--/"~/ath'shaped Fibrous
200 boehmite . ~ / boehmite boehmite
~ / Spherical kaolinite
/ Spherical kaolinite at h-shaped/ Fibrous Fibrous
I-. 150 boehmite / boehmite boehmite
Halloysfe
Low High
Si- CONC_.,EffRATION Figure 9. Schematic diagram showing relationship between formation condition and corresponding minerals. Italic char- acters indicate metastable phases.
could be observed in any products. The spheres ini- tially occurred at the surface of obsidian in very small size (mean diameter of 1-day product was 0.09 #m) and increased progressively in size with time, suggest- ing that the kaolinite was formed as a thermodynam- ically stable phase by a process of heterogeneous nu- cleation on the obsidian surface and successive crystal growth in supersaturated solution.
Halloysite appeared together with fibrous boehmite and trace amounts of allophane and spherical kaolinite in the products of only 150~ reaction using 4.0 g of starting material (Table 1). This probably suggests that relatively low temperature and high Si-activity may be required for its formation. The halloysite was rapidly formed by 1-day reaction but no significant changes in both particle size and abundance could be recognized throughout the reaction up to 60 days. The amounts of fibrous boehmite and allophane also remained un- changed as well as halloysite, whereas the spherical kaolinite increased slightly in abundance with time. Thus, it is undoubtful that the halloysite occurred as a metastable phase in this system. The formation mechanism of halloysite remains uncertain in spite of one of the most abundant clay minerals occurring in the earth 's surface (Minato and Utada 1969, Nagasawa 1978a, 1978b, Tazaki 1982, Velde 1985, Murray 1988, Dixon 1989). Parham (1969) and Busenberg (1979) have synthesized halloysite from feldspar by experi- mental alteration in aqueous solutions, and it was pre- sumed that the halloysite was transformed from feld- spar through a transitional phase of allophane. Sudo and Takahashi (1956) carried out electron microscopic study o f some halloysites altered from volcanic glass
Vol. 43, No. 2, 1995 Formation of clay minerals from obsidian in acid solution 221
and postulated the transformation sequence as follows: volcanic glass --~ allophane --* halloysite. They also stated that extremely fine particles of allophane will coagulate into rounded grains at initial stage and will change subsequently into an aggregate of hydrated hal- loysite with a low degree ofcrystallinity having shapes like twisted fibers or curled hairs. Unfortunately, a transitional phase o f allophane to halloysite in this sequence has not been clearly observed, so it is unclear whether halloysite forms by precipitation due to dis- solution of allophane or by solid state transition from allophane. According to the present study, two distinct types of homogeneous and heterogeneous aggregates of allophane were clearly observed by TEM in the prod- ucts of 150~ reaction (Figures 5a and 5b). The former is a characteristic texture of natural allophane (Wada 1989), whereas the later is an unusual form exhibiting aggregated texture of rounded grains composed of fine particles of allophane. Thus, the rounded grains may be a transitional phase from allophane to halloysite and an individual rounded grain may develop pro- gressively to a spherical halloysite by solid state tran- sition mechanism as the reaction proceeds.
A C K N O W L E D G M E N T S
The authors wish to thank T. Kakoi (Kagoshima University) for the technical assistance extended in the transmission electron microscopy, Y. Ozono (Kago- shima University) for the X-ray photoelectron spec- troscopy, and Y. Kamino (Kagoshima Prefectural In- stitute of Industrial Technology) for the energy dis- persive X-ray analyzer. We also would like to thank the staff of the Institute o f Earth Sciences, Faculty of Science, Kagoshima Universi ty for their generosities.
REFERENCES
Allen, B. L., and B. F. Hajek. 1989. Mineral occurrence in soil environments. In Minerals in Soil Environments. J. B. Dixon and S. B. Weed, eds. Wisconsin: Soil Science Society, 199-278.
Brindiey, G. W., and W. W. Radoslovich. 1956. X-ray stud- ies of the alteration of soda feldspar. In Proc. Fourth Nat. Conf. Clays Clay Minerals, Vol. 4. Ada Swineford, ed. Na- tional Academy of Science: Washington, D.C.
Brown, G. 1980. Associated minerals. In Crystal Structure of Clay Minerals and Their X-ray Identification. G. W. Brindley and G. Brown, eds. London: Mineralogical Soci- ety, 361-410.
Busenberg, E. 1978. The products ofthe interaction offeld- spars with aqueous solutions at 25"C. Geochim. Cosmo- chim. Acta 42: 1679-1686.
Calvet, 1~., P. Boivinet, M. No~l, H. Thibon, A. Maillard, and R. Tertian. 1953. Contribution ~ l'6tude des gels d'alu- mine. Bull. Soc. Chim. Ft. 20: 99-108.
Chou, L., and R. Wollast. 1984. Study of the weathering of albite at room temperature and pressure with a fluidized bed reactor. Geochim. Cosmochim. Acta 48: 2205-2218.
Christoph, G. G., C. E. Corbat6, D. A. Hofmann, and R. T. Tettenhorst. 1979. Thecrystalstructureofboehmite. Clays & Clay Miner. 27: 81-86.
Dixon, J.B. 1989. Kaolin and serpentine group minerals. In Minerals in Soil Environments. J. B. Dixon and S. B. Weed, eds. Wisconsin: Soil Science Society of America, 467-525.
Farmer, V. C., A. R. Fraser, and J. M. Tait. 1979. Char- acterization of the chemical structure of natural and syn- thetic aluminosilicate gels and sols by infrared spectros- copy. Geochim. Cosmochim. Acta 43: 1417-1420.
Gruner, J. W. 1944. The hydrothermal alteration on feld- spars in acid solutions between 300* and 400"C. Econ. Geol. 39: 578-589.
Hall, P. L., G. J. Churchman, and B. K. G. Theng. 1985. Size distribution of allophane unit particles in aqueous sus- pensions. Clays & Clay Miner. 33: 345-349.
Helgeson, H. C., R. M. Garrels, and T. Mackenzie. 1969. Evaluation of irreversible reactions in geochemical pro- cesses involving minerals and aqueous solutions--II. Ap- plications. Geochim. Cosmochim. Acta 33:455-481.
Helgeson, H.C. 1971. Kinetics of mass transfer among sil- icates and aqueous solutions. Geochim. Cosmochim. Acta 35: 421-469.
Hinckley, D.N. 1963. Variability in "crystaUinity" values among the kaolin deposits of the coastal plain of Georgia and South Carolina. Clays & Clay Miner. 11: 229-235.
Holdren, G. H. Jr., and R. A. Berner. 1979. Mechanism of feldspar weathering--I. Experimental studies. Geochim. Cosmochim. Acta 43:1161-1171.
Huang, W. H., and W. D. Keller. 1973. New stability dia- grams of some phyllosilicates in the SiO2-A1203-K20-H20 system. Clays & Clay Miner. 21: 331-336.
Kawano, M., and K. Tomita. 1992. Formation ofallophane and beidellite during hydrothermal alteration of volcanic glass below 200~ Clays & Clay Miner. 40: 666-674.
Kawano, M., and K. Tomita. 1993. Formation of clay min- erals during low temperature hydrothermal alteration of obsidian (Part 1). Effect of addition of AI ions. Nendo Ka- gaku (J. Clay Sci. Soc. Japan) 33:59-71 (in Japanese).
Kawano, M., K. Tomita, andY. Kamino. 1993. Formation of clay minerals during low temperature experimental al- teration of obsidian. Clays & Clay Miner. 41:431-441.
Kittrick, J .A. 1970. Precipitation of kaolinite at 25"C and latm. Clays & Clay Miner, lg: 216-267.
La Iglesia, A., and E. Galan. 1975. HaUoysite-kaolinite transformation at room temperature. Clays & Clay Miner. 23:109-113.
La Iglesia, A., J. L. Martin-Vivaldi Jr., and F. Lopez Aguayo. 1976. Kaolinite crystallization at room temperature by homogeneous precipitation--II: Hydrolysis of feldspars. Clays & Clay Miner. 24: 36-42.
La Iglesia, A., and M. C. Van Oosterwyck-Gastuche. 1978. Kaolinite synthesis. I. Crystallization conditions at low temperatures and calculation of thermodynamic equilibria. Application to laboratory and field observations. Clays & Clay Miner. 26: 397-408.
Minato, H., and M. Utada. 1969. Mode of occurrence and mineralogy ofhaUoysite. In Proc. Inter. Clay Conf., Tokyo, 1969, Vol. I. L. Heller, ed. Jerusalem: Israel University Press, 393-402.
Morey, G. W., and W. T. Chen. 1955. The action of hot water on some feldspars. Amer. Mineral. 40: 996-1000.
Morey, G. W., and R. O. Fournier. 1961. The decompo- sition ofmicrocline, albite and nepheline in hot water. Amer. Mineral 46: 688-699.
Murray, H.H. 1988. Kaolin minerals: Their genesis of oc- currences. In Hydrous Phyllosilicates (Exclusive of Micas). S. W. Bailey, ed. Reviews in Mineralogy, Vol. 19. New York: Mineralogical Society of America, 67-89.
Nagasawa, K. 1978a. Weathering of volcanic ash and other pyroclastic materials. In Clays & Clay Minerals of Japan.
222 Kawano and Tomita Clays and Clay Minerals
T. Sudo and S. Shimoda, eds. Developments in Sedimen- tology, 26. Amsterdam: Elsevier, 105-125.
Nagasawa, K. 1978b. Kaolin minerals. In Clays & Clay Minerals of Japan. T. Sudo and S. Shimoda, eds. Devel- opments in Sedimentology, 26. Amsterdam: Elsevier, 189- 219.
Pap6e, D., R. Tertian, and R. Biais. 1958. Recherches sur la constitution des gels et des hydrates cristallis6s d'alumine. Bull. Soc. Chim. Fr., Mere. Ser. 5: 1301-1310.
Parfitt, R. L., and J. M. Kimble. 1989. Conditions for for- marion ofallophane in soils. Soil Sci. Soc. Am. J. 53 :971- 977.
Parham, W.E. 1969. Formation ofhalloysite from feldspar. Low temperature artificial weathering versus natural weath- ering. Clays & Clay Miner. 17: 13-22.
Robbie, R. A., and D. R. Waldbaum. 1968. Thermodynam- ic properties of minerals and related substances at 298.15*K (25.0"C) and one atmosphere (1.013 bars) pressure and at higher temperatures. U.S. Geol. Surv. Bull. 1259:256 pp.
Roy, R., and E. F. Osborn. 1954. The system A1203-SiO2- H~O. Amer. Mineral. 39: 853-885.
Siefferann, G., and G. Millot. 1969. Equatorial and tropical weathering of Recent basalts from Cameron: Allophane, halloysite, metahalloysite, kaolinite and gibbsite. In Proc. Inter. Clay Conf., Tokyo, 1969, Vol. 1. L. Heller, ed. Je- rusalem: Israel University Press, 417-430.
Sudo, T., and H. Takahashi. 1956. Shapes of halloysite particles in Japanese clays. Clays & Clay Miner. 4: 67-79.
Sudo, T., and H. Takahashi. 197t. The chlorites and inter- stratified minerals. In The Electron-optical Investigation of Clays. J. A. Gard, ed. London: Mineralogical Society, 277- 300.
Tamura, T., and M. L. Jackson. 1953. Structural and energy relationships in the formation of iron and aluminium ox- ides, hydroxides and silicates. Science 117:381-383.
Tazaki, K. 1978. Micromorphology of plagioclase surface at incipient stage of weathering. Earth Science (Chikyu Ka- gaku) 32 :58-62 (in Japanese).
Tazaki, K. 1982. Analytical electron microscopic studies of halloysite formation processes--Morphology and compo- sition of halloysite. In Inter. Clay Conf., 1981. H. van O1- phen and F. Veniale, eds. Developments in Sedimentology, 35. Amsterdam: Elsevier, 573-584.
Tchoubar, C. 1965. Formation de la kaolinite a partir d'al-
bite alt6r6e par l 'eau ~ 200~ Etude en microscopie et dif- fraction 61ectroniques. Bull. Soc. Franf. Min~r. Crist. 88: 483-518.
Tettenhorst, R., and D. A. Hofmann. 1980. Crystal chem- istry of boehmite. Clays & Clay Miner. 28: 373-380.
Thomas, F.B. 1971. The kaolin minerals. In The Electron- optical Investigation of Clays. J. A. Gard, ed. London: Min- eralogical Society, 109-157.
Tomura, S., Y. Shibasaki, H. Mizuta, and M. Kitamura. 1983. Spherical kaolinite: Synthesis and mineralogical properties. Clays & Clay Miner. 31: 413-421.
Tomura, S., Y. Shibasaki, H. Mizuta, and M. Kitamura. 1985. Growth conditions and genesis of spherical and platy ka- olinite. Clays & Clay Miner. 33: 200-206.
Tsuzuki, Y., and K. Suzuki. 1980. Experimental study of the alteration process of labradorite in acid hydrothermal solutions. Geochim. Cosmochim. Acta 44: 673--683.
Tsuzuki, Y., and I. Kawabe. 1983. Polymorphic transfor- mations of kaolin minerals in aqueous solutions. Geochim. Cosmochim. Acta 47: 59-66.
van Olphen, H. 1971. Amorphous clay materials. Science 171: 90-91.
Velde, B. 1985. Clay Minerals, A physico-Chemical Expla- nation of their Occurrence, Developments in Sedimentology, 40: Amsterdam: Elsevier, 427 pp.
Wada, K. I989. Allophane and imogolite. In Minerals in Soil Environments. J. B. Dixon and S. B. Weed, eds. Wis- consin: Soil Science Society of America, 1051-1087.
Wada, S.-I., A. Eto, and K. Wada. 1979. Synthetic allophane and imogolite. J. Soil Sci. 30: 347-355.
Wada, S.-I., and K. Wada. 1981. Formation between alu- minate ions and orthosilicic acid in dilute alkaline to neutral solutions. Soil Sci. 132: 267-273.
Watanabe, T., and T. Sudo. t969. Study on small-angle scattering of some clay minerals. In Proc. Int. Clay Conf., Tokyo, 1969, Vol. 1. L. Heller, ed. Jerusalem: Israel Uni- versity Press, 173-181.
Wells, N., C. W. Childs, and C. J. Downes. 1977. Silica spring, Tongariro National Park, New Zealand--Analysis of the spring water and characterization of the alumino- silicate deposit. Geochim. Cosmochim. Acta 41: 1498-1506.
(Received 18 February 1994; accepted 21 September 1994; Ms. 2478)
